Exploring Explainable AI and Reinforcement Learning in Educational Conversational Agents with Metacognitive Self-Regulated Learning: A Systematic Literature Review
International Journal of Innovative Computing · 28.07.2026 · 16 (1-2) · 171-177
DOI: 10.11113/ijic.v16n1-2.690 ↗Kurz gefasst
Systematische Übersicht in der Zeitschrift International Journal of Innovative Computing, online erschienen am 28.07.2026. Verfasst von Li Haoze und 5 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This systematic literature review explores integrating conversational AI, reinforcement learning (RL), and interpretable AI (XAI) in programming education to enhance metacognitive self-regulated learning (MSRL). Conversational AI offers programming support and feedback, and combining it with RL optimizes interaction strategies, yet existing studies suffer from overreliance on generative AI, insufficient long-term empirical evidence, and AI "black box" issues. XAI, while addressing the black box problem, struggles to balance performance and interpretability. Core research questions include effective RL integration for MSRL, XRL ’ s role in building trust, conversational AI’s support for code generation and metacognitive training, application challenges, and long-term experiment design. A "conversational AI+RL+XAI+affective computing" framework was established via database searches, with mixed methods analyzing quantitative and qualitative data. Findings reveal a focus on paired technology combinations rather than overall synergy. Key challenges are mismatched XAI interpretation strategies, difficulty quantifying MSRL as RL rewards, and lack of standardized evaluation. The study fills critical research gaps, providing a foundation for transparent adaptive educational AI. Future research should prioritize process interpretability, large-scale long-term teaching experiments, and optimized multi-objective reward functions to advance programming education tutoring systems.
Zitieren
APA
Haoze, L., Ibrahim, N., Hoe, C. W., Shafie, M. L. M., Halim, S. A., & Saadon, N. A. (2026). Exploring Explainable AI and Reinforcement Learning in Educational Conversational Agents with Metacognitive Self-Regulated Learning: A Systematic Literature Review. International Journal of Innovative Computing, 16(1-2), 171-177. https://doi.org/10.11113/ijic.v16n1-2.690
Vancouver
Haoze L, Ibrahim N, Hoe CW, Shafie MLM, Halim SA, Saadon NA. Exploring Explainable AI and Reinforcement Learning in Educational Conversational Agents with Metacognitive Self-Regulated Learning: A Systematic Literature Review. International Journal of Innovative Computing. 2026;16(1-2):171-177. doi:10.11113/ijic.v16n1-2.690
BibTeX
@article{Haoze2026,
author = {Haoze, Li and Ibrahim, Noraini and Hoe, Chan Weng and Shafie, Muhammad Luqman Mohd and Halim, Shahliza Abd and Saadon, Nor Azizah},
title = {Exploring Explainable AI and Reinforcement Learning in Educational Conversational Agents with Metacognitive Self-Regulated Learning: A Systematic Literature Review},
journal = {International Journal of Innovative Computing},
year = {2026},
volume = {16},
number = {1-2},
pages = {171--177},
doi = {10.11113/ijic.v16n1-2.690}
}
EUVIMED